[Paper Review] OpenCV2X: Modelling of the V2X Cellular Sidelink and Performance Evaluation for Aperiodic Traffic

Brian J. McCarthy, Andres Burbano-Abril+2 more|arXiv (Cornell University)|Mar 24, 2021

Advanced MIMO Systems Optimization25 citations

TL;DR

OpenCV2X is the first open-source full-stack simulator of 3GPP Release 14 C-V2X sidelink, used to evaluate SB-SPS under aperiodic CAM traffic and explore dedicated aperiodic scheduling mechanisms.

ABSTRACT

This paper presents OpenCV2X, the first publicly available, open-source simulation model of the Third Generation Partnership Project (3GPP) Release 14 Cellular Vehicle to Everything (C-V2X) sidelink, which forms the basis for 5G NR Mode 2 under later releases. This model is fully compliant with the existing vehicular service and application layers, including messaging sets as defined by the automotive and standards communities providing a fully standardised, cross-layer communication model. Using this model, we show how the current sidelink scheduling mechanism performs poorly when scheduling applications with highly aperiodic communication characteristics, such as ETSI Cooperative Awareness Messages (CAMs). We then provide the first indepth evaluation of dedicated per-packet aperiodic scheduling mechanisms, in contrast to schemes that parameterise the existing algorithm. This paper highlights that the level of aperiodicity exhibited by the application model greatly impacts scheduling performance. Finally, we analyse how such scheduling mechanisms might co-exist.

Motivation & Objective

• Provide a full-stack, open-source C-V2X sidelink simulator compliant with ETSI/3GPP messaging and cross-layer standards.
• Characterise SB-SPS performance with highly aperiodic traffic patterns such as CAMs.
• Evaluate dedicated aperiodic scheduling mechanisms and co-existence with SB-SPS for mixed traffic.
• Inform design choices for NR V2X Mode 2/Mode 4 prioritisation and future 5G NR-based V2X.
• Assess conditions under which SB-SPS degrades with aperiodic traffic.

Proposed method

• Develop a full-stack, open-source simulator (OpenCV2X) extending SimuLTE and integrating Artery/INET or Veins for cross-layer V2X modeling.
• Implement C-V2X Mode 4 PHY/MAC layers (LtePhyVueMode4, LteMacVueMode4) and support SB-SPS with sensing/CSR filtering.
• Distribute power across transmitted Resource Blocks to compute PSSCH-RSRP and S-RSSI per RB for accurate BLER/packet error rate calculations.
• Validate OpenCV2X against an analytical LTE-V Mode 4 model and compare Channel Busy Ratio (CBR) and packet loss attribution.
• Model ETSI CAMs and 3GPP aperiodic traffic, then test dedicated aperiodic scheduling mechanisms and mixed traffic coexistence.

Experimental results

Research questions

• RQ1 How does SB-SPS perform when faced with highly aperiodic application traffic (e.g., CAMs) under varying vehicular densities?
• RQ2 What conditions cause SB-SPS performance degradation for aperiodic traffic, and can dedicated aperiodic scheduling mechanisms mitigate it?
• RQ3 Do dedicated aperiodic scheduling mechanisms co-exist effectively with SB-SPS for mixed application models?
• RQ4 Which application models (3GPP vs ETSI CAMs) dominate channel load and influence PDR under different transmit powers and MCSs?
• RQ5 How do changes in transmission power and MCS affect PDR and interference in V2X Mode 4?

Key findings

• SB-SPS performance degrades with highly aperiodic traffic, primarily due to frequent grant breaks and increased collisions.
• Dedicated per-packet aperiodic scheduling mechanisms can significantly improve performance for highly aperiodic traffic, while less variable inter-arrival rates can be accommodated by parameter adjustments in SB-SPS.
• Application models with higher aperiodicity (ETSI CAMs) exhibit notably poorer PDR, especially as density increases, compared to more periodic patterns.
• Higher transmission power and higher MCS can improve PDR but trade off robustness; adaptive MCS/subchannel configurations may help reduce congestion.
• OpenCV2X validation shows PDR and CBR closely align with analytical LTE-V Mode 4 models, with mean absolute deviations generally below a few percent across densities.